Fresh of the press: Preliminary Mayan results … [Norcliffe & Jaeger]

The orange juice is still warm, the cafe con helado barely melted, the Mexican music that has been playing on repeat for the last couple of hours still swings mind-numbingly in my fried brain (it’s VERY hot and humid here), and here we are: letting you, dear reader [sic], know what the world is waiting for: is there probability-sensitive morphosyntactic production in Yucatec Mayan (similar to English, cf. Frank & Jaeger, 2008-CUNY, 2008-CogSci; Jaeger, 2006-thesis, 207-LSA; Levy & Jaeger, 2007; Wasow et al., in press)?

Getting started in Mexico: Contacts & Pilots on Mayan

As some of you know, we’ve been planning to study certain aspect of language production in Mayan for some time now. Well, planning has been followed by flying, and now we (Elisabeth Norcliffe, Stanford University, and I) are here and ready to run our first studies!

CogSci08 - here we come

I still don’t know how I managed to not ever have been to CogSci before, but this year it will happen. Thanks to Austin Frank (BCS, UofR), Carlos Gomez Gallo (CS, UofR), and Neal Snider (Ling, Stanford University), a bunch of us will be presenting at CogSci08 in Washington, D.C. in July. I will upload the papers soon.

What’s up?

Phew, time for a short update before I am off again.

• Celeste Kidd just gave her first conference talk at Prosody08 on the pronunciation of function words, investigating the combined effects of availability-based production, redundancy avoidance, collocations, and prosody on the pronunciation of the English indefinite determiner “a”.
• Carlos Gomez Gallo is about to present his work on a multi-modal corpus (video, sound, and plenty of linguistically useful annotation) at LREC08 in Marrakesh, Morocco. This work builds the foundation to Carlos’s CogSci paper (to be presented in July) on planning in language production above the level of the clause [paper].
• Katrina Housel and Ting Qian are about to give there presentation at the University of Rochester Undergraduate Research Workshop. Katrina will talk about her work on the effect of phonological similarity between words in spontaneous sentence production; Ting will talk about his work on Constant Entropy in written Chinese and written English by Chinese speakers.

CUNY was fun

That’s all. I mean I had fun. Oh, and before there are more questions. The stats program mentioned during one of the sessions (for multilevel or mixed models) is R, which is freely available and you can learn more about it by following the R-lang list on the right of this blog. R is shell-based, which means you will need some time to get used to it, but it is a powerful program with implementations of most of the types of models that everyone is talking about these days and there is already a rather large community of R-users interested in language research (see the R-lang email list). Enjoy.

The mysterious .L, .Q., and .C

Ok, I know this is dumb of me (obviously), but I bet I am not the only one ever who wondered what those mysterious x.L, x.Q., and x.C variables are that R automatically creates when an ordered factor x is entered into a model (lm, glm, lmer, etc.).  Well, it should be kinda obvious, but with enough whiskey those things can stand for anything.

Anyway, on the odd chance that somebody will google for those terms and look for salvation: The default coding R uses for ordered predictors is simply polynomial contrast coding … (Linear, Quadratic, Cubic, etc.) .

Modeling self-paced reading data: Effects of word length, word position, spill-over, etc.

I’ve been using a two-step approach, where in the first step I use all data (including fillers, but not practice items) of an experiment to fit a model of log-transformed raw reading times with:

• word length (Wlen)
• position of word in stimulus (Wpos)
• position of stimulus in list (Lpos)

R-code for reading time data preparation

Some time ago I posted some R-code on how to create spill-over data from a linger reading time file (for spill-over analysis of self-paced reading time data). Here are the steps that need to be done prior to that, importing from a linger file, data preparation, outlier check, etc.

I wish I could vote

Should you or should you not? Why, of course, you should …

leave comments and feedback of the sensible and/or entertaining kinda (preferably related to the posts).

one thanks. two, too.

Florian